JP2013092917A - Health management supporting device, health management supporting method and computer program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a health management supporting device which makes a management supporting method for efficiently maintaining a healthy body possible, and further to provide a health management supporting method and a computer program.SOLUTION: A health management supporting device comprises: multi-band image acquiring sections 12, 32 and 50; a spectral estimation parameter storage section 36; a calibration processing parameter storage section 38; a calibration processing section 18; a diagnostic database storage section 39; a nutritional management database storage section 40 and an exercise database storage section 42; a user information storage section 44; and a diagnosing section 19 which calculates an evaluation value of an object to be tested T from a feature amount obtained by the calibration processing section 18 by using diagnostic data stored in the diagnostic database storage section 39 and calculates an exercise and an exercise amount from the evaluation value of the object to be tested T by using nutritional management data stored in the nutritional management database storage section 40 and exercise data stored in the exercise database storage section 42.

Description

  The present invention relates to a health management support device, a health management support method, and a computer program.

  In recent years, lifestyle-related diseases such as hypertension, diabetes, and hyperlipidemia have become a problem with changes in the living environment. Lifestyle-related diseases tend to forget that daily meals are important during a busy life, and dietary habits such as insufficient vegetable intake and excessive intake of salt and fat are considered as one of the factors. Therefore, in general, it is said that it is important for lifestyle-related disease prevention to review daily lifestyle habits, eat moderately exercised, and have a well-balanced diet that can provide the nutrients that should be taken a day.

  Conventionally, a plurality of meal menus are stored in advance, a good menu is proposed from personal information (age, sex, height, weight, etc.), and the food eaten from the proposed menu is selected and input to identify calories. Moreover, the calorie consumption of various exercise | movement is memorize | stored previously, the meal menu of the day and the exercise amount of the day are input, and the difference of calorie consumption is displayed from the total intake calorie. Furthermore, a health management system has been proposed in which exercise and exercise amount for consumption of excess calories are displayed and the quality of the exercise consists of walking (see, for example, Patent Document 1).

JP 2006-251871 A

  However, in Patent Document 1, selection from a registered meal menu is fundamental, and it is not possible to deal with unregistered meals. In addition, since the calorie amount is registered in advance, it cannot cope with the calorie fluctuation of the meal amount (large, large eaten, left over). Furthermore, since the calories of the eaten food are not directly measured, it cannot cope with fluctuations due to differences in ingredients and cooking methods. Furthermore, the management of the day is fundamental, but the management of “the excess and deficiency of the day” generated by the schedule and physical condition is also important. Also, exercise other than walking cannot be selected.

  SUMMARY An advantage of some aspects of the invention is to solve at least a part of the problems described above, and the invention can be implemented as the following forms or application examples.

  Application Example 1 The health management support apparatus according to this application example acquires a multiband image obtained by imaging the subject in a plurality of wavelength bands including at least a predetermined wavelength band corresponding to a specific component of the subject. A multiband image acquisition unit that performs the spectral estimation parameter storage unit that stores a spectral estimation parameter used to convert a multiband image having a different wavelength band into a spectral spectrum, and a feature amount relating to a specific component of the subject. A spectral processing parameter storage unit that stores calibration processing parameters used for conversion into a spectral processing parameter, and a spectral calculation that calculates the spectral spectrum from the multiband image using the spectral estimation parameter stored in the spectral estimation parameter storage unit An estimation unit and the calibration processing parameter stored in the calibration processing parameter storage unit. A calibration processing unit that calculates the feature amount from the spectral spectrum obtained by the spectral estimation unit using a meter, and diagnostic data used to convert the feature amount of the subject into an evaluation value of the subject A diagnostic database storage unit that stores the nutrition management data storage unit that stores nutrition management data used to convert the evaluation value of the subject into exercise and exercise amount, and an exercise database storage unit that stores exercise data, The feature information obtained by the calibration processing unit using the user information storage unit that stores user information used for referring to the nutrition management data and the exercise data, and the diagnostic data stored in the diagnostic database storage unit The evaluation value of the subject is calculated from the nutrition, and the nutrition stored in the nutrition management database storage unit Using the movement data stored in the sense data and the motion data base storage unit, the characterized in that it comprises a diagnosis unit, a for calculating the movement amount and the movement from the evaluation value of the subject.

  According to this application example, the function of the food calorie measuring instrument is utilized to propose exercise and exercise amount that eliminate the excess excluding basal metabolism from the amount of food consumed. In the case of calorie deficiency, suggest a meal amount to compensate for the deficiency. In addition, by proposing the amount of meal before and after and “appropriate amount of exercise” for “I want to eat OO!”, Management of the amount of meal (INPUT) and the amount of exercise (OUTPUT) can be easily continued (continue), As a result, a health management support device that enables a management support method for efficiently maintaining a healthy body is provided.

  Application Example 2 In the health management support apparatus according to the application example, the diagnosis unit outputs the health management to a display unit.

  According to this application example, health management can be presented easily.

  Application Example 3 In the health management support apparatus according to the application example described above, the multiband image acquisition unit instructs a measurement band to a multiband camera that captures the subject with sensitivity of a plurality of wavelength bands. A measurement band data storage unit for storing a plurality of measurement band data used for the measurement band, and sending the measurement band data stored in the measurement band data storage unit to the multiband camera. A measurement band instruction unit for instructing.

  According to this application example, since it is possible to instruct the measurement band to the multiband camera, a multiband image of a plurality of wavelength bands including a predetermined band corresponding to the specific component can be more reliably acquired.

  Application Example 4 In the health management support apparatus according to the application example described above, a correction amount used for correcting an error in a multiband camera from which the multiband image is acquired is a band image that forms the multiband image. A preprocessing data storage unit for storing preprocessing data determined for each of the first and second multiband images acquired by the multiband image acquisition unit based on the preprocessing data stored in the preprocessing data storage unit. And a pre-processing unit that transmits the corrected multiband image to the spectral estimation unit.

  According to this application example, since the error in the multiband camera can be corrected, the estimation accuracy can be further improved.

  Application Example 5 The health management support method according to this application example obtains a multiband image obtained by imaging the subject in a plurality of wavelength bands including at least a predetermined wavelength band corresponding to a specific component of the subject. Using the spectral estimation parameter used to convert the multiband image having a different wavelength band into a spectral spectrum, calculating the spectral spectrum from the multiband image, and analyzing the spectral spectrum of the subject. A step of calculating the feature amount from the spectral spectrum obtained by the calculation using a calibration processing parameter used for conversion to a feature amount relating to a specific component, and converting the feature amount into an evaluation value of the subject Calculating the evaluation value of the subject from the feature amount obtained by the calculation using diagnostic data used for A step of calculating the exercise and the amount of exercise from the evaluation value of the subject obtained by the calculation using nutrition management data and exercise data used for converting the evaluation value of the subject into exercise and exercise amount. It is characterized by including these.

  According to this application example, the feature amount of the subject can be estimated with high accuracy from the multiband image without requiring an expensive spectroscope.

  Application Example 6 A computer program according to this application example obtains a multiband image obtained by imaging the subject in a plurality of wavelength bands including at least a predetermined wavelength band corresponding to a specific component of the subject. And a function for calculating the spectral spectrum from the multiband image using spectral estimation parameters used for converting the multiband image having different wavelength bands into a spectral spectrum, and the spectral spectrum as a specific component of the subject. A function for calculating the feature amount from the spectral spectrum obtained by the calculation using a calibration processing parameter used for converting the feature amount into a feature amount, and converting the feature amount into an evaluation value of the subject The evaluation value of the subject is calculated from the feature amount obtained by the calculation using diagnostic data to be used Using the nutrition management data and exercise data used to convert the evaluation value of the subject into exercise and exercise amount, and the exercise and the exercise amount from the evaluation value of the subject obtained by the calculation. It is characterized by having a computer realize the function of calculating

  According to this application example, the feature amount of the subject can be estimated with high accuracy from the multiband image without requiring an expensive spectroscope.

  Furthermore, the present invention can be realized in various forms other than the first to sixth application examples. For example, a form of a health management support system including the health management support apparatus according to the first to fourth application examples together with a multiband camera. Can be realized.

Explanatory drawing which shows schematically the structure of the health management assistance system which concerns on a present Example. The flowchart which shows the health care support system which concerns on a present Example. The figure which shows the screen transition in the initial screen which concerns on a present Example. The figure which shows the screen transition in the determination screen which concerns on a present Example. The figure which shows the screen transition in the determination recording screen and carry-over confirmation screen concerning a present Example. The figure which shows the screen transition in the exercise | movement confirmation screen and weight reduction confirmation screen which concern on a present Example. The table | surface which shows the activity content of the physical activity level which concerns on a present Example. The flowchart which shows the flow of determination of the ingestible amount which concerns on a present Example OK. The flowchart in which the determination of the ingestible amount which concerns on a present Example shows the flow of NG. The table | surface which shows the calorie list consumed by the exercise | movement which concerns on a present Example. The flowchart which shows the exercise navigation which concerns on a present Example. The figure which shows the screen of the exercise navigation which concerns on a present Example. The table | surface which shows an exercise ranking with the high calorie consumption which concerns on a present Example. The flowchart which shows imaging | photography of the dish which concerns on a present Example. The flowchart which shows the production | generation / preservation | save of the database which concerns on a present Example. The table | surface which shows the nutrition management database which concerns on a present Example. The table | surface which shows the exercise | movement database which concerns on a present Example. The flowchart which shows the spectral image process which concerns on a present Example.

  Hereinafter, the present embodiment will be described based on examples. The present embodiment is for estimating a feature amount related to a specific component of a meal that is a subject and supporting health management.

A. Overall structure of health management support system:
FIG. 1 is an explanatory diagram schematically illustrating a configuration of a health management support system according to the present embodiment. As illustrated, the health management support system 1 includes a spectral image processing apparatus 100, a multiband camera 200, a display device (display unit) 300, and a storage device 400. The multiband camera 200, the display device 300, and the storage device 400 are electrically connected to the spectral image processing device 100.

  The multiband camera 200 includes a lens unit 210, a wavelength selection filter 220, a CCD 230, a CCD AFE (Analog Front End) 240, a light source unit 250, and the like. The lens unit 210 does not include an autofocus mechanism that focuses the subject T, but may include an autofocus mechanism. The wavelength selection filter 220 is a Fabry-Perot filter that can change the transmission wavelength band. The CCD 230 is an imaging device that photoelectrically converts light that has passed through the wavelength selection filter 220 to obtain an electrical signal representing a subject T image. The CCD AFE 240 is for digitizing the detection signal of the CCD 230. The light source unit 250 is for irradiating the subject T.

  In the multiband camera 200 having the above-described configuration, the wavelength selection filter 220 sequentially receives a plurality of measurement band instructions from the spectral image processing apparatus 100, whereby the transmission wavelength range of the wavelength selection filter 220 is sequentially changed. In this way, the multiband camera 200 images the subject T with sensitivity in a plurality of wavelength bands (multiband).

  The display device 300 is a device for displaying information on a screen. The storage device 400 is an external device for storing data, and is, for example, a hard disk drive device.

  The spectral image processing apparatus 100 is an apparatus that determines the freshness of the subject T by processing a multiband image obtained by photographing with the multiband camera 200. The spectral image processing apparatus 100 receives image data from a CPU 10 that performs various processes and controls by executing a program, a memory 30 that stores programs, data, and information, and a multiband camera 200, and a display apparatus 300. And an input / output interface (I / F) 50 for sending the result of diagnosis to the storage device 400.

  The memory 30 includes a measurement band data storage unit 32, a preprocessing data storage unit 34, a spectral estimation parameter storage unit 36, a calibration processing parameter storage unit 38, a diagnostic database storage unit 39, a nutrition management database storage unit 40, and an exercise database storage unit 42. And a user information storage unit 44. Details of each of the storage units 32 to 44 will be described later. The memory 30 includes a program storage unit that stores a health management support program (not shown). In this embodiment, the memory 30 is a non-volatile memory. In addition, it is good also as a structure which receives and preserve | saves data and parameters required before execution of a process as a volatile memory instead of a non-volatile.

  The CPU 10 executes the health management support program stored in the program storage unit of the memory 30, thereby allowing the measurement band instruction unit 12, the preprocessing unit 14, the spectral estimation unit 16, the calibration processing unit 18, and the diagnosis unit 19. Realize functionally. Each part 12-19 performs each process using the data and parameter preserve | saved at the corresponding preservation | save parts 32-44. The measurement band instruction unit 12 instructs the multiband camera 200 to specify the measurement band by sending the measurement band data stored in the measurement band data storage unit 32 to the multiband camera 200. The preprocessing unit 14 corrects the multiband image acquired by the measurement band instruction unit 12 based on the preprocessing data stored in the preprocess data storage unit 34, and converts the corrected multiband image into the spectral estimation unit 16. Send to. The spectral estimation unit 16 calculates a spectral spectrum from the multiband image using the spectral estimation parameters stored in the spectral estimation parameter storage unit 36. The calibration processing unit 18 calculates a feature amount from the spectral spectrum obtained by the spectral estimation unit 16 using the calibration processing parameter stored in the calibration processing parameter storage unit 38. The diagnosis unit 19 calculates the evaluation value of the subject T from the feature amount obtained by the calibration processing unit 18 using the diagnosis database (diagnosis data) stored in the diagnosis database storage unit 39. The diagnosis unit 19 calculates exercise and exercise amount from the evaluation value of the subject T using the nutrition management data stored in the nutrition management database storage unit 40 and the exercise data stored in the exercise database storage unit 42. The diagnosis unit 19 uses the user information for initial value calculation of the nutrition management database (nutrition management data) and reference (gender) of the exercise database (exercise data). The diagnosis unit 19 outputs health management to the display device 300. As a result of each processing, the CPU 10 acquires a multiband image obtained by the multiband camera 200, supports health management of the subject T from the multiband image, and displays the search result via the input / output I / F 50. The data is transmitted to the device 300 and the storage device 400. As a result, the result of health management support is displayed.

  Note that the spectral image processing apparatus 100 also functions as a feature quantity estimation apparatus because it estimates a feature quantity related to a specific component of the subject T when supporting the health management by the subject T. The input / output interface (I / F) 50, the measurement band data storage unit 32 of the memory 30, and the measurement band instruction unit 12 that functions in the CPU 10 correspond to the “multiband image acquisition unit” in Application Example 1.

B. Work procedures for health care support:
FIG. 2 is a flowchart illustrating the health management support system according to the present embodiment. FIG. 3 is a diagram illustrating screen transition in the initial screen according to the present embodiment. FIG. 4 is a diagram illustrating screen transition in the determination screen according to the present embodiment. FIG. 5 is a diagram illustrating screen transitions in the determination recording screen and the carry-over confirmation screen according to the present embodiment. FIG. 6 is a diagram illustrating screen transitions on the exercise confirmation screen and the weight loss confirmation screen according to the present embodiment.

  First, in step S100, the CPU 10 calculates a daily required nutrient amount (= Z). Calculated from Harris-Benedix equation (HBE) (basal metabolism) physical activity level. At that time, the CPU 10 uses the user information data of the user information storage unit 44.

  As a formula for calculating basal metabolism (basic energy consumption: BEE), there is a Harris-Benedict formula (HBE). The daily energy consumption varies from person to person and depends on factors such as gender, weight, and age. HBE is a mathematical formula used to calculate the amount of energy (kcal / day) required by a healthy person in a resting state, and is represented by the following formula. Male: BEE = 66.4730 + 13.7516w + 5.30033h−6.7550a, Female: BEE = 655.955 + 9.5634w + 1.8496h−1.8496a (w: body weight (kg), h: height (cm), a: age ( (Years old)), HBE is a somewhat complicated formula, so there is a simple formula created to calculate Japanese BEE. Male: BEE = 14.1w + 620, Female: BEE = 10.8w + 620.

  The BEE for calculating the required energy amount is the minimum energy amount necessary for survival. In order to calculate the energy actually required, it is necessary to calculate by multiplying values called activity coefficients and stress coefficients (invasion factors). In addition, HBE is a calculation formula originally created for Europeans and Americans, so when applied to Japanese, it may be calculated slightly higher than the actual value.

  The nutrition according to the present embodiment is the daily energy requirement (kcal) = the daily basal metabolic rate (kcal) × the physical activity level. For example, if you are in your 30s and are commuting to work desk-centered (= physical activity level is normal, that is, the value is 1.75), the daily energy requirement is 1520 kcal. 1520 (kcal) × 1.75 = 2660 (kcal). In a woman with a basal metabolic rate of 1140 kcal, it is 1140 (kcal) × 1.75 = 1995 kcal).

  FIG. 7 is a table showing activity contents at physical activity levels according to the present embodiment. The physical activity level is a representative value, and the range in () is an approximate range. In parentheses in each activity classification (time / day) is an activity factor (Af value indicating the intensity per unit time in each physical activity, expressed as a multiple of basal metabolism) (representative value: lower limit to upper limit). .

  Next, in step S110, the CPU 10 reads the morning / daytime / evening section. For example, the CPU 10 may determine the morning / daytime / evening classification from the operation time and the number of operations.

  Next, in step S120, the CPU 10 reads the carry-over nutrition (Y). The carry-over nutrient amount (= Y) is the carry-over amount of the previous day. The carry-over nutrient amount (Y) is a plus amount and a minus amount, and the updated carry-over nutrient amount (Y ′) is read after the second meal.

  Next, in step S <b> 130, the CPU 10 calculates the acquired nutrient amount average value (X) for morning, noon, and evening. The CPU 10 calculates the acquired nutrient average value (X) from the morning, noon, and evening categories.

  Next, in step S140, the CPU 10 photographs a dish (see C for details).

  Next, in step S150, the CPU 10 reads the nutritional calibration (V) on the “calibration screen” shown in FIG.

  Next, in step S160, the CPU 10 determines an ingestible amount. If X ≦ V + Y (OK in step S160), the process proceeds to step S170.

  Next, in step S170, the CPU 10 turns on the green lamp on the “determination screen: OK” shown in FIG. Then, the process proceeds to step S200.

  If X> V + Y (NG in step S160), the process proceeds to step S180.

  Next, in step S180, the CPU 10 lights the yellow lamp on the “determination screen: NG” shown in FIG. Then, the process proceeds to step S300.

FIG. 8 is a flowchart showing a flow of OK determination of the ingestible amount according to the present embodiment.
First, in step S200, the CPU 10 proposes a meal that supplements the deficient nutrients. The CPU 10 suggests what and how much should be consumed as a diet to supplement deficient nutrients.

  In this embodiment, the amount of exercise is proposed based on what is currently eaten and the amount of meal before and after that. For example, the average calorie value for every three meals is recorded from past data.

  In this embodiment, since long-term dietary restrictions cause stress, the amount of meal before and after and the amount of exercise are proposed from the object (purpose) to be eaten. If it cannot be resolved before and after, control it on a weekly basis. For example, the average calorie value for every three meals is recorded from past data. Specifically, if you eat a lot during lunch, the actual calorie amount of breakfast, the expected calorie amount of lunch, and the average calorie value of dinner "Exercise" is presented.

  In this embodiment, in the case of deficiency, a meal is proposed that takes into account the balance of lipids, carbohydrates, and proteins from the taste of meal history. For example, food images may be recorded and uploaded to Blog along with the calorie content.

  Next, in step S210, the CPU 10 determines whether or not there is ingestion. When not ingesting (No in step S210), the process proceeds to step S220. If ingested (Yes in step S210), the process proceeds to step S230.

  Next, in step S220, the CPU 10 calculates the intake nutrient amount (V ′ = V + Y). Then, the process proceeds to step S250.

  Next, in step S230, the CPU 10 reads a nutritional calibration (R).

  Next, in step S240, the CPU 10 calculates an intake nutrition amount (V ′ = V + Y + R). Then, the process proceeds to step S250.

  Next, in step S250, the CPU 10 calculates the acquired nutrient average value (X) of the nutrient intake (morning, noon, evening) V ′.

  Next, in step S260, the CPU 10 calculates a carry-over nutrient amount (Y ′ = V′−X). In that case, you may display a carry-over nutrient amount like the "carry-over recording screen" shown in FIG. This makes me feel happy to eat a little more in the next meal.

  Next, in step S270, the CPU 10 displays the main level (Z ′ = Z−V ′) of daily nutrition as shown in FIG. The intake level Z ′ with respect to the daily required nutrient amount Z is displayed. For example, it may be displayed as a bar graph. In the present embodiment, the calorie excess that could not be settled on that day is carried over from the next day.

FIG. 9 is a flowchart showing the flow of NG when determining the ingestible amount according to the present embodiment.
First, in step S300, the CPU 10 determines whether exercise can be performed. When exercise is performed (Yes in step S300), the process proceeds to step S310. When not exercising (No in step S300), the process proceeds to step S350.

  Next, in step S310, the CPU 10 displays the recommended exercise amount as in the “exercise selection screen” shown in FIG. The CPU 10 displays and selects several types of candidates (exercise content, time).

  Next, in step S320, the CPU 10 reads the execution result on the “exercise confirmation screen” shown in FIG.

  Next, in step S330, the CPU 10 calculates the exercise calorie consumption (U).

  FIG. 10 is a table showing a list of calories consumed in the exercise according to the present embodiment. For example, the CPU 10 may calculate the amount of calories consumed (U) from the calories consumed by the exercise shown in FIG.

  Next, in step S340, the CPU 10 calculates the intake nutrient amount (V ′ = V + Y−U). Then, the process proceeds to step S250.

  Next, in step S350, the CPU 10 determines whether to reduce or settle the next meal. When decreasing (decrease in step S350), the process proceeds to step S360. In the case of liquidation at the next meal (clearing at step S350), the process proceeds to step S400. For example, sometimes there are things that you really want to eat and sometimes you want to eat a lot.

  Next, in step S360, the CPU 10 displays an amount that can be eaten (X ≦ V + Y). The CPU 10 may display a numerical value indicating what percentage of the photographed meal image is used as a guide.

  Next, in step S370, the CPU 10 photographs the remaining food (see C for details).

  Next, in step S380, the CPU 10 reads a nutritional calibration (R).

  Next, in step S390, the CPU 10 calculates the intake nutrient amount (V ′ = V + Y−R). Then, the process proceeds to step S250.

  Next, in step S400, the CPU 10 calculates the intake nutrient amount (V ′ = V + Y). Then, the process proceeds to step S260.

FIG. 11 is a flowchart showing the exercise navigation according to the present embodiment. FIG. 12 is a diagram illustrating an exercise navigation screen according to the present embodiment.
First, in step S500, the CPU 10 reads the carry-over calorie amount.

  Next, in Step S510, the CPU 10 reads the selection result of the extra time on the “extra time selection screen” shown in FIG. For example, 1.0H is read.

  Next, in step S520, the CPU 10 displays the recommended exercise amount according to the remaining display time on the “recommended exercise amount screen” shown in FIG. The CPU 10 displays and selects several types of candidates (exercise content, time). In this embodiment, an appropriate exercise is proposed from the input extra time.

  FIG. 13 is a table showing an exercise ranking with a high calorie consumption according to the present embodiment. The calorie consumption is a value when each exercise is performed for 1 hour. For example, the CPU 10 may display the recommended exercise amount from the exercise ranking with the high calorie consumption shown in FIG.

C. Cooking procedure:
FIG. 14 is a flowchart illustrating the shooting of food according to the present embodiment. As shown in the figure, first, in step S <b> 600, the CPU 10 generates a database necessary for cooking photography, and stores the database in the memory 30 of the spectral image processing apparatus 100. The database referred to here corresponds to various data stored in the storage units 32 to 44 (see FIG. 1) in the spectral image processing apparatus 100 (see C-1 for details).

  Next, in step S <b> 630, the CPU 10 images the subject T using the multiband camera 200 connected to the spectral image processing apparatus 100. The multiband camera 200 irradiates the subject T with near-infrared light from the light source unit 250 and images the reflected light with the multiband camera 200. For the measurement wavelength band, the measurement band instruction unit 12 reads the measurement band data stored in the measurement band data storage unit 32 and controls the wavelength selection filter 220 of the multiband camera 200. The illumination for photographing has absorption of sucrose, lipid, protein and water necessary for obtaining calories, and near infrared light (600 to 2500 nm) in which light reaches a relatively deep part of the sample is preferable. Note that the light source unit 250 is not necessarily included in the multiband camera 200. In addition, if the measurement band data matched to the calibration / diagnosis target is generated in the generation / storage of the database in step S600, the process of photographing the meal image in step S630 performs a common process regardless of the calibration / diagnosis target. Can be used.

  Next, in step S640, the CPU 10 estimates the spectral spectrum / feature amount of each pixel from the multiband image output from the multiband camera 200 connected to the spectral image processing apparatus 100, and obtains the calorie of the dish. The amount / distribution (diagnosis result) of carbohydrates, lipids, proteins, and water necessary for the determination is obtained (refer to C-2 for details).

  Next, in step S <b> 660, the CPU 10 displays the feature amount and the diagnosis result on the display device 300 connected to the spectral image processing apparatus 100.

  Next, in step S670, the CPU 10 determines whether or not there is a next subject T. If there is a next subject T (Yes in step S670), the process returns to the meal menu search in step S630, and steps S630 to S660 are performed on the next subject T. In step S670, if there is no next subject T (No in step S670), the process is terminated.

C-1. Database creation and storage:
FIG. 15 is a flowchart illustrating database generation / storage according to the present embodiment. The contents and generation method of various data as the database generated and stored in step S600 will be described in detail below. The various data are measurement band data, preprocessing data, spectral estimation parameters, calibration processing parameters, a diagnostic database, and usage history. Note that it is preferable to generate and save the database before the spectral image processing apparatus 100 is shipped from the factory. The determination in step S640 may be omitted when the subject T whose diagnosis result is known is used. Part of the above procedure may be performed by an operator through an interface (not shown).

  First, in step S602, the CPU 10 generates and stores preprocess data. The multiband camera 200 has hardware manufacturing errors such as optical system unevenness, warpage of a Fabry-Perot filter constituting the wavelength selection filter 220, and illumination unevenness of the light source unit 250. Accordingly, in the multiband camera 200, wavelength unevenness (transmission wavelength unevenness) and light amount unevenness occur in each plane of the multiband image obtained by photographing (in each band image constituting the multiband image). In this embodiment, the correction amount for canceling the in-plane wavelength unevenness and the light amount unevenness is stored in the preprocess data storage unit 34 as preprocess data. The preprocess data storage unit 34 determines the correction amount used for correcting the error in the multiband camera 200 that is the acquisition destination of the multiband image for each of the band images constituting the multiband image. Save.

  Specifically, first, a sample of uniform color is photographed with a multiband camera 200 (the same model, more preferably the actual machine itself) used for health care support, and a multiband image is acquired. Next, in each of the band images constituting the multiband image, correction data for canceling the hardware manufacturing error is obtained for each pixel so as to be uniform at any position in the image, and the correction data for each pixel is obtained. The set is stored in the preprocess data storage unit 34 as preprocess data corresponding to each band image. The uniform color sample is preferably a color that can reflect light evenly in any wavelength range. For example, a standard white plate is prepared.

  The preprocess data is a table of correction amounts for each pixel of the multiband image. The correction amount is obtained by measuring in advance unevenness in the optical system / filter of the multiband camera 200 and illumination unevenness in the light source unit 250 and calculating a value for canceling the measurement. Since the preprocess data depends only on the characteristics of the multiband camera 200, the process of generating and storing the preprocess data in step S602 can use a common process regardless of the calibration / diagnosis target.

  Next, in step S604, the CPU 10 generates and stores a spectral estimation parameter and measurement band data. Spectral estimation parameters are used in spectral spectrum estimation. The spectral estimation parameter is stored in the spectral estimation parameter storage unit 36. The spectral estimation parameter storage unit 36 stores spectral estimation parameters used to convert multiband images having different wavelength bands into spectral spectra.

  The measured bandwidth data is used for searching the meal menu. The measurement band data is stored in the measurement band data storage unit 32. The measurement band data storage unit 32 stores a plurality of measurement band data used to instruct the measurement band to the multiband camera 200 that captures an image of the subject T with sensitivity of a plurality of wavelength bands.

  The spectral estimation parameter represents the relationship between the multiband image and the spectral spectrum. The multiband image is a plurality of band images having different wavelength bands obtained by the multiband camera 200, and the spectral spectrum is a spectrum obtained by a spectroscope. Specifically, the spectral estimation parameter represents the relationship between the pixel value x of the multiband image and the spectral spectrum p, and corresponds to a spectral estimation matrix M in the following equation.

p = Mx, x = ^t (x ₁ , x ₂ ,... x _n ), p = ^t (p ₁ , p ₂ ,..., p _m )
The spectral estimation matrix M is calculated using, for example, the following equation.

^t M = (X ^t X) ⁻¹ X ^t P, X = (x ₁ , x ₂ ,... x _l ), P = (p ₁ , p ₂ ,... p _l )
Here, X is a pixel value of a multiband image in a plurality of spectral estimation samples, and P is also a spectral spectrum.

  The measurement band data is obtained by searching for a condition that can maintain the spectral estimation accuracy with a small number of bands while changing the measurement band and the number of bands of the multiband value x when calculating the spectral estimation parameter.

  Note that the spectral estimation parameter can be changed to content other than the above as long as it can represent the relationship between the multiband image and the spectral spectrum. Also, the method for generating the spectral estimation parameter need not be limited to the above, and can be replaced with another method. Furthermore, the measurement band data generation process need not be limited to the above, and can be replaced with other methods such as principal component analysis. A color chart or the like may be used as the spectral estimation sample. As a result, the spectral estimation parameter becomes a value that depends only on the multiband camera 200, and thus can be used in common regardless of the subject T and the feature amount.

  Next, in step S606, the CPU 10 generates and stores calibration processing parameters. Calibration processing parameters are used in spectral estimation. The calibration processing parameters are stored in the calibration processing parameter storage unit 38. The calibration processing parameter storage unit 38 stores calibration processing parameters used for converting a spectral spectrum into a feature amount. The calibration processing parameter represents the relationship between the spectral spectrum and the characteristic amount of the subject T. Specifically, the calibration processing parameter represents the relationship between the spectral spectrum p and the feature quantity s, and corresponds to the mixing matrix A in the following equation.

s = Ap, p = ^t (p ₁ , p ₂ ,... p _m )
The mixing matrix A is calculated using, for example, the following formula.

^t A = (P ^t P) ⁻¹ P ^t s, P = (p ₁ , p ₂ ,..., p _l ), s = (s ₁ , s ₂ ,..., _sl )
Here, P is a spectral spectrum in a plurality of calibration samples, and s is a feature quantity. Note that the target dish may be used as the calibration sample. In addition, for each sample, a spectral spectrum in the same wavelength band as the multiband image may be measured with a spectrophotometer, and the amount of nutrients (such as carbohydrates, lipids, proteins, and water) may be measured by chemical analysis. . Furthermore, the processing for obtaining the calibration processing parameter from the spectral spectrum and the feature amount may use a common method regardless of the calibration / diagnosis target. Further, the calibration processing parameter can be changed to contents other than the above as long as it can express the relationship between the spectral spectrum and the characteristic amount of the subject T. Furthermore, the method for generating calibration processing parameters is not limited to the above, and can be replaced with other methods such as a method using principal component analysis, PLS analysis, neural network, or the like.

  Next, in step S608, the CPU 10 generates and stores a diagnostic database. The diagnostic database is used to estimate the amount of calibrated material. The diagnostic database is stored in the diagnostic database storage unit 39. The diagnostic database storage unit 39 stores a diagnostic database used for converting the feature amount of the subject T into an evaluation value of the subject T. The diagnosis database is a table of diagnosis results corresponding to each feature amount. For example, the diagnostic database includes daily target calorie intake / nutrients, age / sex / weight, and the like. Daily target calorie intake / nutrients are determined according to age / sex / weight. In addition, the user first inputs his / her age / sex / weight.

  FIG. 16 is a table showing a nutrition management database according to the present embodiment. The nutrition management database storage unit 40 stores nutrition management data used for converting the evaluation value of the subject T into exercise and exercise amount. The unit of each numerical value in the nutrition management database of the nutrition management database storage unit 40 is kcal. The initial value of the nutrition management database is used as the average value of nutrition at the start of system use. For example, for males, BEE = 66.4730 + 13.7516w + 5.30033h−6.7550a (w: weight (kg), h: height (cm), a: age (years)), BEE = 66.4730 + (13. 7526 * 67) + (5.00033 * 175)-(6.7550 * 43) = 1573 kcal (weight 67 kg, height 175 cm, age 43 years), daily energy requirement (kcal) = daily basal metabolism The quantity (BEEkcal) × physical activity level is 1573 kcal * 1.4 = 2200 kcal (physical activity level 1.4). There are various theories regarding the amount of calorie distribution in the morning, day, and night, but in this embodiment, the distribution is adjusted slightly based on the distribution suitable for diet, morning: noon: night = 2: 3: 1. The acquired nutrient amount average value is an average intake calorie value of each meal (every morning and night) in the past month.

  FIG. 17 is a table showing an exercise database according to the present embodiment. The exercise database storage unit 42 stores exercise data used for converting the evaluation value of the subject T into exercise and exercise amount. The exercise database of the exercise database storage unit 42 indicates, for example, calorie consumption per hour for each exercise.

  The user information storage unit 44 stores user information used for referring to the nutrition management database and the exercise database. The user information of the user information storage unit 44 is used for initial value calculation of the nutrition management database and reference (gender) of the exercise database. Each data referred to is used by the diagnosis unit 19 for determination and processing. The user information includes, for example, parameters such as height / weight / age / sex / life intensity (physical activity level, occupation type) / pregnancy. The setting items of the user information include, for example, time setting, use / non-use of the carry-on nutrition of the previous day on the next day, physical activity level setting for the day (can be changed as necessary), diet mode, and the like. If the previous day's nutrition is not carried over to the next day, reset the date when the date is changed. You can change your physical activity level on the day you exercise or work hard. In diet mode, the total nutrition is operated at a low value.

C-2. Spectral image processing:
FIG. 18 is a flowchart illustrating spectral image processing according to the present embodiment. Details of each process are shown below.
First, in step S642, the CPU 10 performs processing for acquiring a multiband image from the multiband camera 200. Specifically, the CPU 10 sequentially transmits a plurality of measurement band data stored in the measurement band data storage unit 32 of the memory 30 to the wavelength selection filter 220 of the multiband camera 200, thereby transmitting the transmission wavelength band of the wavelength selection filter 220. Are sequentially changed, and multiband images representing band images (spectral images) for different measurement bands are acquired from the multiband camera 200. In addition, the structure which transmits several measurement band data to the multiband camera 200 which is a part of process in this step S642 corresponds to the measurement band instruction | indication part 12 (FIG. 1) implement | achieved by CPU10.

  Next, in step S644, the CPU 10 executes preprocessing for correcting each of the band images acquired in step S642. The multiband image includes measurement errors caused by unevenness of the optical system / filter of the multiband camera 200 and illumination unevenness of the light source unit 250. This is corrected using the preprocess data. The processing in step S644 corresponds to the preprocessing unit 14 (FIG. 1) realized by the CPU 10.

  Next, in step S646, the CPU 10 executes processing for estimating a spectral spectrum from the multiband image after correction in step S644 (hereinafter referred to as “corrected multiband image”). The corrected multiband image obtained by correcting the unevenness of the camera / light source in step S644 is a discrete measurement value, and is a value integrated in the transmission wavelength band of the filter of the multiband camera 200. There is a difference from the spectrum of the reflected light from. This is estimated by the following formula using the spectral estimation parameter in step S604 and the spectral estimation parameter generated and stored in the generation and storage of the measurement band data.

p = Mx, x = ^t (x ₁ , x ₂ ,... x _n ), p = ^t (p ₁ , p ₂ ,..., p _m )
Note that, similarly to the generation and storage of the spectral estimation parameter and measurement band data in step S604, the spectral estimation processing is not limited to the above, and can be replaced with another method.

  Next, in step S648, the CPU 10 performs a process of estimating the feature amount of the subject T, that is, the calorie / nutrient amount, from the spectral spectrum obtained in step S646. Here, the estimation is performed based on the calibration processing parameters stored in the calibration processing parameter storage unit 38.

  Specifically, using the following equation from the spectral spectrum of each pixel obtained in the spectral spectrum estimation in step S646 and the obtained calibration processing parameter generated and stored in the calibration processing parameter generation and storage in step S606 Estimate the amount.

s = Ap, p = ^t (p ₁ , p ₂ ,... p _m )
Note that, similarly to the generation and storage of the calibration processing parameters in step S606, the calibration processing is not limited to the above, and can be replaced with another method.

  The processing in step S646 corresponds to the spectral estimation unit 16 (FIG. 1) realized by the CPU 10, and the processing in step S648 corresponds to the calibration processing unit 18 (FIG. 1) realized by the CPU 10.

Next, in step S650, the CPU 10 performs processing for obtaining a diagnosis result from the feature amount obtained in step S648. Here, the freshness determination result is calculated based on the diagnostic database stored in the diagnostic database storage unit 39.
The calculated calorie / nutrient of the meal is added to the calorie / nutrient amount already consumed on the same day, and compared with the daily target calorie / nutrient amount stored in the diagnosis database storage unit 39 to diagnose Find the result. Further, the diagnosis result is obtained based on the degree of deviation from the target value.

  Next, in step S652, the CPU 10 outputs the diagnosis result obtained in step S650 to the display device 300. Display of the diagnostic result obtained by calculating the diagnostic result in step S650, the feature quantity obtained by estimating the calibration substance amount in step S648, the spectral spectrum obtained by estimating the spectral spectrum in step S646, the diagnosis date and time, location information, etc. Output to the device 300.

  In addition, if the spectral estimation parameter and the calibration processing parameter matched to the calibration / diagnosis target are generated in the generation and storage of the database in step S600 shown in FIG. 14, the spectral image processing in step S640 does not depend on the calibration / diagnosis target. Common processing can be used.

  According to this embodiment, the minimum necessary number of measurement bands are measured according to the substance to be measured, and the spectral spectrum is estimated and calibrated from the data. Therefore, it is possible to achieve both measurement time and accuracy. It can be measured.

  In addition, since the spectral spectrum can be estimated even if the spectral sensitivity range of the sensor is wide, the amount of light incident on the sensor increases, and even higher resolution sensors and inexpensive sensors with poor S / N can be measured. It can measure at low cost and resolution.

  In addition, meal menus that match the daily target calorie intake / nutrient differences, time, place, and other conditions are presented, so you can choose meals that you can actually eat on the spot, and the trouble of selecting meal menus By omitting, it is possible to save the trouble of selecting a meal menu.

  According to this embodiment, the calorie content of the actual meal can be measured from the spectral information. Thereby, there is little error of the calorie amount (it is not a replacement of the preset calorie amount). Further, it is not necessary to register the menu of the system and create a meal database, and the amount of memory of the system can be reduced.

  In addition, green (OK) or yellow (NG) lamps are turned on according to the amount of meal OK or NG, or the entire display means is green or yellow. Thereby, it is possible to measure (input) quickly on the spot.

  Furthermore, the result of the nutritional calibration is output as the total calorie amount and PFC balance (protein, lipid, carbohydrate). As a result, the calorie amount can be easily input (output), so that the user can continue to use it.

  In addition, the average value of acquired nutrients is calculated from past meal records (the accuracy increases with the number of uses). The first time is about morning: daytime: night = 3: 4.5: 2.5. Thereby, the meal according to a user's preference and real life is possible.

  Furthermore, it is possible to set a calorie balance (morning: noon: evening intake ratio) of three meals. For example, it is possible to make settings such as conserving the daytime and focusing on the daytime. This enables positive health management for the purpose of eating.

D. Variations:
The present invention is not limited to the above-described embodiments and modifications thereof, and can be carried out in various modes without departing from the gist thereof. For example, the following modifications are possible. In the description of the modification, the same components as those in the above-described embodiment are denoted by the same reference numerals, and the description thereof is omitted.

(Modification 1)
In the above embodiment, the spectral image processing apparatus 100 dedicated to the health management support system 1 is used. However, the spectral image processing apparatus 100 can be replaced with a general-purpose personal computer. Moreover, although the multiband camera 200 and the spectral image processing apparatus 100 are separate bodies, they can be integrated. For example, the spectral image processing apparatus 100 may be built in, or the multiband camera 200 may be built with the spectral image processing apparatus 100 in the form of a chip, for example. The display device 300 and the storage device 400 can also be integrated with the spectral image processing device 100.

(Modification 2)
In the embodiment and each modified example, the function realized by software may be realized by hardware.

(Modification 3)
In the embodiment and each of the modified examples, the spectral spectrum is calculated from the multiband image using the spectral estimation parameter, and the feature amount is calculated from the spectral spectrum using the calibration processing parameter. Thus, the feature quantity may be directly estimated from the multiband image using the parameter L (= Z · M) obtained by multiplying the spectral estimation parameter M and the calibration processing parameter Z. When the combination of the individual of the multiband camera 200, the type of the subject T, and the feature amount to be estimated is determined in advance, this configuration also allows the subject to be detected from the multiband image without requiring an expensive spectroscope. The feature amount of the specimen can be estimated with high accuracy.

(Modification 4)
In the embodiment and each modification, the transmission wavelength band of the wavelength selection filter 220 is transmitted by sequentially transmitting the plurality of measurement band data stored in the measurement band data storage unit 32 to the wavelength selection filter 220 of the multiband camera 200. Are sequentially changed to obtain a multiband image representing a band image (spectral image) for each different measurement band from the multiband camera 200. Instead of this, a multi-band image representing a band image for each of a plurality of predetermined measurement bands may be acquired from the multi-band camera 200. That is, the measurement band data storage unit 32 and the measurement band instruction unit 12 may be omitted.

(Modification 5)
In the embodiment and each modification, the preprocessing unit 14 corrects the multiband image based on the preprocessing data stored in the preprocessing data storage unit 34, and estimates the spectral spectrum from the corrected multiband image; did. Instead of this, a configuration may be adopted in which the spectral spectrum is estimated from the multiband image acquired by the multiband image acquisition unit. Further, the multiband image acquisition unit may acquire a corrected multiband image. That is, the preprocessing data storage unit 34 and the preprocessing unit 14 may be omitted.

(Modification 6)
In the embodiment and each modification, the diagnosis unit 19 makes a determination based on the diagnosis database stored in the diagnosis database storage unit 39 of the spectral image processing apparatus 100 and the estimated feature amount. Instead, the estimated feature amount may be output to a diagnostic device configured separately from the spectral image processing device 100, and the diagnostic device may perform diagnosis. That is, the spectral image processing apparatus 100 may not include the diagnostic database storage unit 39 and the diagnostic unit 19.

(Modification 7)
In the embodiment, the green lamp is turned on in step S170 shown in FIG. In step S180, the yellow lamp is turned on. Instead, it may be identified by sound, vibration or the like. Further, only OK / NG display may be used.

(Modification 8)
In the said Example, it was set as the structure which proposes the meal which supplements a deficient nutrient by step S200, S210, S230, S240 shown in FIG. Instead of this, the proposal of a diet that supplements the deficient nutrients may be omitted.

  It should be noted that elements other than the elements described in the independent claims among the constituent elements in the above-described embodiments and modifications are additional elements and can be omitted as appropriate.

  DESCRIPTION OF SYMBOLS 1 ... Health management support system 10 ... CPU 12 ... Measurement band instruction | indication part (multiband image acquisition part) 14 ... Pre-processing part 16 ... Spectral estimation part 18 ... Calibration processing part 19 ... Diagnosis part 30 ... Memory 32 ... Measurement band data preservation | save Unit (multiband image acquisition unit) 34 ... preprocessing data storage unit 36 ... spectral estimation parameter storage unit 38 ... calibration processing parameter storage unit 39 ... diagnosis database storage unit 40 ... nutrition management database storage unit 42 ... exercise database storage unit 44 ... User information storage unit 50 ... Input / output I / F (multiband image acquisition unit) 100 ... Spectral image processing device 200 ... Multiband camera 210 ... Lens unit 220 ... Wavelength selection filter 230 ... CCD 240 ... CCDAFE 250 ... Light source unit 300 ... Display device 400 ... memory device .

Claims (6)

A multiband image acquisition unit for acquiring a multiband image obtained by imaging the subject in a plurality of wavelength bands including at least a predetermined wavelength band corresponding to a specific component of the subject;
A spectral estimation parameter storage unit that stores spectral estimation parameters used to convert the multiband images having different wavelength bands into spectral spectra;
A calibration processing parameter storage unit for storing a calibration processing parameter used for converting the spectral spectrum into a characteristic amount related to a specific component of the subject;
A spectral estimation unit that calculates the spectral spectrum from the multiband image using the spectral estimation parameter stored in the spectral estimation parameter storage unit;
A calibration processing unit that calculates the feature amount from the spectral spectrum obtained by the spectral estimation unit using the calibration processing parameter stored in the calibration processing parameter storage unit;
A diagnostic database storage unit for storing diagnostic data used to convert the feature quantity of the subject into an evaluation value of the subject;
A nutrition management database storage unit that stores nutrition management data used to convert the evaluation value of the subject into exercise and exercise amount; and an exercise database storage unit that stores exercise data;
A user information storage unit for storing user information used for referring to the nutrition management data and the exercise data;
Using the diagnostic data stored in the diagnostic database storage unit, calculating the evaluation value of the subject from the feature amount obtained by the calibration processing unit,
Diagnosis of calculating the exercise and the amount of exercise from the evaluation value of the subject using the nutrition management data stored in the nutrition management database storage unit and the exercise data stored in the exercise database storage unit And
A health management support apparatus comprising: The health management support apparatus according to claim 1,
The health diagnosis support apparatus, wherein the diagnosis unit outputs the health management to a display unit. In the health care support apparatus according to claim 1 or 2,
The multiband image acquisition unit
A measurement band data storage unit for storing a plurality of measurement band data used for instructing a measurement band to a multiband camera that performs imaging of the subject with sensitivity of a plurality of wavelength bands;
A measurement band instruction unit for instructing the measurement band to the multiband camera by sending the measurement band data stored in the measurement band data storage unit to the multiband camera;
A health management support apparatus comprising: In the health care support apparatus according to any one of claims 1 to 3,
Pre-processing data storage for storing pre-processing data in which a correction amount used to correct an error in a multi-band camera from which the multi-band image is acquired is determined for each of the band images constituting the multi-band image And
A pre-processing unit that corrects the multi-band image acquired by the multi-band image acquisition unit based on pre-processing data stored in the pre-processing data storage unit, and sends the corrected multi-band image to the spectral estimation unit When,
A health management support apparatus comprising: Obtaining a multiband image obtained by imaging the subject in a plurality of wavelength bands including at least a predetermined wavelength band corresponding to a specific component of the subject; and
Calculating the spectral spectrum from the multiband image using spectral estimation parameters used to convert the multiband image having different wavelength bands into a spectral spectrum;
A step of calculating the feature amount from the spectrum obtained by the calculation using a calibration processing parameter used for converting the spectral spectrum into a feature amount relating to a specific component of the subject;
Calculating the evaluation value of the subject from the feature amount obtained by the calculation using diagnostic data used to convert the feature amount into an evaluation value of the subject;
A step of calculating the exercise and the amount of exercise from the evaluation value of the subject obtained by the calculation using nutrition management data and exercise data used for converting the evaluation value of the subject into exercise and exercise amount. When,
A health care support method comprising: A function of acquiring a multiband image obtained by imaging the subject in a plurality of wavelength bands including at least a predetermined wavelength band corresponding to a specific component of the subject;
A function of calculating the spectral spectrum from the multiband image using spectral estimation parameters used to convert the multiband image having different wavelength bands into a spectral spectrum;
A function of calculating the feature value from the spectrum obtained by the calculation using a calibration processing parameter used for converting the spectrum into a feature value related to a specific component of the subject;
A function of calculating the evaluation value of the subject from the feature amount obtained by the calculation using diagnostic data used to convert the feature amount into an evaluation value of the subject;
A function of calculating the exercise and the amount of exercise from the evaluation value of the subject obtained by the calculation using the nutrition management data and the exercise data used for converting the evaluation value of the subject into exercise and the amount of exercise. When,
A computer program characterized in that a computer is realized.